JPH07182455A - Character recognition device/method - Google Patents

Abstract

PURPOSE:To quality the input characters without performing any special operation in accordance with the writing input of characters. CONSTITUTION:This character recognition device/method is provided with a coordinate input device 11 which contains the writing frames 51a-51f and acquires a coordinate value string of characters to these writing frames, a character recognition dictionary 41 where the character information corresponding to the coordinate value string is previously registered, and the arithmetic processing executing parts 21-25 which decide whether the coordinate value string acquired by the device 11 overflows the frames 51a-51f by a fixed extent or more, recognize the characters inputted based on the dictionary 41 by the coordinate value string, output the character qualification information after adding this information to the characters recognized by the result of the preceding decision, and display the outputted character information at the position of a display device 13 corresponding to the frames 51a-51f of the device 11 in accordance with the presence or absence of the character qualification information.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a character recognition device and method for recognizing a character input by handwriting through a coordinate input device such as a tablet.

[0002]

2. Description of the Related Art In recent years, attention has been paid to an online character recognition device for handwriting and recognizing characters through a coordinate input device such as a tablet and recognizing the characters for document creation. In a general device of this type, a writing frame is displayed in advance on a display screen configured integrally with a tablet, and characters are input and recognized one by one so that the writing frame fits within the writing frame. ing. Therefore, it is not possible to write extra information different from characters in the writing frame.

[0003]

Therefore, in the character recognition device as described above, when the writer wants to input a modified character such as a double-width character, the desired character is input in the writing frame. In addition to the above, it is necessary to perform an operation such as inputting an instruction button or icon with a pen tip for modification operation for character editing displayed at another position on the display screen.

As described above, in order to obtain a character with some modification, it is necessary to perform an operation which has nothing to do with the act of writing a character during the input, so that the handwriting can be efficiently input. In addition to being unable to do so, there was a problem that the writer's thoughts could be disturbed.

The present invention has been made in view of the above circumstances, and an object of the present invention is to apply a modification to an input character without performing a special operation accompanying the handwriting input of the character. To provide a character recognition device and method capable of

[0006]

That is, the present invention has a writing frame, a coordinate input device for obtaining a coordinate value sequence of a character handwritten in the writing frame, and character information corresponding to the coordinate value sequence. And a recognition dictionary in which the coordinate value sequence obtained by the coordinate input device exceeds the writing frame by a certain amount or more, and the recognition is performed by the coordinate value sequence obtained by the coordinate input device. While recognizing the characters input in the writing frame of the coordinate inputting means by referring to the dictionary and adding the character modification information to the characters recognized by the judgment result and outputting the output character information. The character pattern is displayed at a position corresponding to the writing frame of the coordinate input device according to the presence or absence of the character.

[0007]

With the above-mentioned structure, it is possible to modify the input character without performing a special operation accompanying the handwriting of the character.

[0008]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS An embodiment of the present invention will be described below with reference to the drawings. FIG. 1 is a diagram showing a schematic circuit configuration of a character recognition device realized by incorporating the device of the embodiment, 11 is a coordinate input device such as a transparent tablet, 12 is a control device which is the device body, and 13 is a liquid crystal display. And the like.

This device has, for example, a structure in which a coordinate input device 11 and a display device 13 are laminated and integrated as shown in FIG. 2 and mounted on a control device 12. Coordinate input device here
The transparent tablet 11 is for detecting and inputting information of a position designated by a pen or the like as coordinate values on the tablet. Further, the liquid crystal display which is the display device 13 has a display surface having the same dimensions and coordinates as the tablet surface, and the display information can be visually confirmed through the transparent tablet.

When the display information is instructed by the transparent tablet, the information of the instructed position is input through the tablet, and the control device 12 will be described later.
The display information can be specified.

As shown in FIG. 1, the control device 12 includes an arithmetic processing execution unit including an input unit 21, a recognition data processing unit 22, a recognition processing unit 23, a display data processing unit 24, and a display processing unit 25.
A buffer memory including a position data buffer 31, a recognition data buffer 32, a display data buffer 33, a character data buffer 34, a character recognition dictionary 41, and a character pattern dictionary 42.
The dictionary memory is composed of.

The input unit 21 processes the coordinate values designated and input from the coordinate input device 11 and sequentially stores the series of coordinate values in the position data buffer 31. Also input section 21
When the icon for inputting various control commands is displayed on the display device 13, it is determined whether or not the input coordinate value indicates the icon, and if the icon is indicated. In accordance with the input command, each operation of the arithmetic processing unit described above is activated.

The recognition data processing unit 22 analyzes the series of coordinate values indicating the input pattern stored in the position data buffer 31 as described above, and obtains the direction value indicating the direction of the writing stroke. Further, the recognition data processing unit 22 detects when a handwriting input is carried out beyond a certain fixed amount from one of the writing frames of the coordinate input device 11 which will be described later, and detects it in an internal flag register (not shown). It is supposed to set a flag.

The calculation of the stroke direction value by the analysis of the coordinate value series performed by the recognition data processing unit 22 is performed by, for example, obtaining the direction of change from the first coordinate value of the input pattern to the next coordinate value and quantizing the direction. It is carried out by determining the direction value corresponding to the converted value. Then, the sequence of the direction values obtained up to the end point coordinates of the input pattern is compressed, for example, by grouping those having the same direction value consecutively, and the sequence of the direction values left minimally is used to form the input pattern. This is performed by using a direction value series. In other words, each time the direction value changes, the direction value is sequentially extracted as information indicating the direction of the stroke.

In this way, the series of direction values obtained by the recognition data processing unit 22 are sequentially stored in the recognition data buffer 32, and are used for pattern recognition processing by the recognition processing unit 23. Then, the recognition processing unit 23 refers to the direction value series of the input pattern stored in the recognition data buffer 32 by referring to the direction value series of the standard pattern registered in advance in the character pattern dictionary 42, and performs the recognition processing as follows. is doing.

That is, in the character recognition dictionary 41, the information of the standard pattern is associated with the character code (the category of the standard pattern) and the series of the direction values thereof, and the number of strokes (the direction value forming the series of the direction values). Are registered in advance.

When the recognition data buffer 32 stores a direction value series relating to one input pattern, the recognition processing section 23 checks the number of direction values forming the direction value series to find a corresponding stroke from the character recognition dictionary 41. The numbers are read serially in order. Then, between the standard pattern and the input pattern, the difference in the direction value is calculated for each corresponding stroke. Then, the recognition processing unit 23 obtains the sum of the difference between the direction values of the strokes obtained as described above.

The above processing is sequentially executed for the standard patterns registered in the character recognition dictionary 41 and having the same number of strokes as the input pattern, and the standard pattern having the smallest sum of the difference of the direction values is recognized for the input pattern. I'm asking as a result.

At this time, a standard pattern in which the sum of the difference between the direction values is smaller than a predetermined value may be obtained as a next recognition candidate so that a correct recognition result can be selected. The input pattern information (character code) recognized in this way is sequentially stored in the character data buffer 34.

Up to this point, it has been described that the character to be input is recognized by directly writing the character to be input into the coordinate input device 11 which is integrally formed on the screen of the display device 13. Displays a writing frame for writing characters on the screen of the display device 13 as shown in FIG. 3, and if it is a normal character input according to this display, the writing frame fits within each writing frame. Write down above.

FIG. 3 shows a plurality of, for example, six writing frames 51a to 51f displayed on the screen of the display device 13.
By displaying the plurality of writing frames 51a to 51f in this manner, it becomes possible to continuously input a plurality of characters by writing. The division (segmentation) between characters is one character when it is judged that the input frame moves to another and the obtained coordinate value sequence is completely discontinuous from the adjacent frame. The recognition data processing unit 22 detects that the input has been completed. Also, considering that there is no frame to write next, there is no writing input for a certain set time,
In addition, when it is determined that the pen is apart from the input surface of the coordinate input device 11 by a certain distance or more and is in the air, the display of the writing frame is erased and the next writing input is awaited.

The recognition data processing unit 22, the recognition processing unit 23, the display data processing unit 24, and the display processing unit 25 manage the writing input and the display of the recognition result while the writing frames 51a to 51f are displayed. Be done.

In the recognition data processing section 22, the writing frames 51a-51 are provided.
The frame in which the handwriting input is performed is managed, the stroke data obtained as a result of the analysis is recognized by the recognition processing unit 23 as described above, and the obtained character code is stored in the character data buffer 34.

Therefore, the display processing unit 25 obtains the standard pattern of the corresponding character from the character pattern dictionary 42 according to the character code obtained as described above. Then, according to the input position of the input pattern, the obtained standard pattern is expanded to the character pattern at a position on the display data buffer 33 corresponding to the frame of the pattern input position. By expanding the character pattern on the display data buffer 33, the character pattern is displayed at the corresponding display position on the display device 13.

Next, a specific operation of this apparatus will be described. FIG. 4 shows a state in which the hiragana “A” is written across two writing frames 51a and 51b for the purpose of double-width character modification. The processing procedure is shown in FIGS.

This processing procedure will be briefly described. In FIG. 6, the recognition data processing unit 22 first clears an internal flag register (shown as "F" in the drawing) to "0" at the beginning of processing (step). A1).

Thereafter, as shown in FIG. 4, when a handwriting input is made across the two writing frames 51a and 51b, the input unit 21 detects the input and sequentially outputs the series of coordinate values as position data. The data is sequentially stored in the buffer 31 (step A2). The recognition data processing unit 22 analyzes the series of coordinate values stored in the position data buffer 31, sequentially stores the series of stroke direction values in the recognition data buffer 32, and determines whether or not there is protrusion from one writing frame. Judge whether or not (step A3
).

If it is determined that the writing frame is present, the recognition data processing unit 22 determines whether the width of the next writing frame exceeds the width of the writing frame by α times (eg, α = 1/3) or more. That is, it is judged whether or not the protrusion is an intentional protrusion exceeding a preset range (step A4).

Then, as shown in FIG. 4, when it is judged that the protruding width is α times or more, the recognition data processing unit 22
Sets "1" in the internal flag register (step A5), then detects the direction of protrusion, in this case, the protrusion from the writing frame 51a to the writing frame 51b (step A6). After temporarily storing the information indicating that, the process ends.

Further, if it is determined in step A3 that there is no protrusion from one writing frame, or if there is protrusion, it is determined in step A4 that the width does not exceed α times the recognition data. The processing unit 22 once ends this processing without performing the flag setting and other processing in steps A5 and A6.

In response to the operation of the recognition data processing unit 22 as described above, the recognition processing unit 23 determines the character pattern dictionary 42 from the direction value series of the input pattern stored in the recognition data buffer 32.
Of the standard pattern, the standard pattern having the smallest sum of the differences is obtained as a recognition result, and the character code thereof is stored in the character data buffer 34.

On the other hand, the display processing unit 25 obtains the standard pattern of the corresponding character from the character pattern dictionary 42 according to the character code stored in the character data buffer 34, and
As shown in FIG. 7, it is determined whether "1" is set in the internal flag register of the recognition data processing unit 22 at that time (step B1).

When it is determined that "1" is set, the display processing unit 25 reads the information indicating the direction protruding from the recognition data processing unit 22 and character-modifies the standard pattern of the character according to the direction. Operate (Step B2
), And then the internal flag register of the recognition data processing unit 22 is cleared to "0" (step B3). In this case, the hiragana “a” standard pattern is expanded and converted in the horizontal direction so as to form a double-width horizontal pattern. And
In accordance with the input position of the writing frame, the obtained character pattern is expanded to the corresponding position on the display data buffer 33, here the positions of the writing frames 51a and 51b, and displayed at the corresponding display position of the display device 13 (step B5) With the above, the processing from the handwriting input to the display of the recognition result is completed.

If it is determined in step B1 that "1" is not set in the internal flag register of the recognition data processing unit 22, the display processing unit 25 uses the standard pattern of the character as it is as a full-width, normal-size font. It will be selected (step B4). Then, according to the input position of the writing frame, a character pattern of a normal size font is developed at the corresponding position on the display data buffer 33 and displayed at the corresponding display position of the display device 13 (step B5
), The processing ends.

Further, instead of writing over the two writing frames as shown in FIG. 4, the writing is performed by protruding the characters from one writing frame toward the upper side as shown in FIG. Can also be considered. FIG. 5 shows a state in which the hiragana “A” is written so as to protrude from one writing frame 51a toward the information for the purpose of performing double-width vertical character modification.

At this time, in FIG. 6, after the recognition data processing unit 22 clears the internal flag register to "0" (step A1), the handwriting input is made as shown in FIG. The input unit 21 detects the input and sequentially stores the series of coordinate values in the position data buffer 31 (step A2). The recognition data processing unit 22 analyzes the coordinate value series stored in the position data buffer 31, sequentially stores the series of stroke direction values in the recognition data buffer 32, and determines whether or not there is a protrusion from one writing frame. Is judged (step A3).

When it is determined that there is a protrusion, the recognition data processing unit 22 next extends the width (height) of the writing frame.
Is .alpha. Times the width of the writing frame (for example, .alpha. = 1/3) or more, that is, it is determined whether or not the intentional protrusion exceeds the preset range (step A4). Here, as shown in FIG. 5, when it is determined that the protruding width (height) is α times or more, the recognition data processing unit 22 sets "1" in the internal flag register (step A5), and then The protruding direction is detected. In this case, it is detected that the writing frame 51a is protruding to the upper side (step A6).
), And temporarily stores information indicating that, and then ends this processing.

In response to the operation of the recognition data processing unit 22 as described above, the recognition processing unit 23 refers to the character pattern dictionary 42 from the direction value series stored in the recognition data buffer 32, and recognizes the hiragana character "a" as a recognition result. Is obtained and the character code is stored in the character data buffer 34.

On the other hand, the display processing unit 25 obtains the standard pattern of the corresponding character from the character pattern dictionary 42 according to the character code of the hiragana "A" stored in the character data buffer 34, and as shown in FIG. At this time, it is determined that "1" is set in the internal flag register of the recognition data processing unit 22 (step B1), the information indicating that the direction protruding from the recognition data processing unit 22 is upward is read, and A character modification operation is performed on the standard pattern of the character according to the direction (step B2), and then the internal flag register of the recognition data processing unit 22 is cleared to "0" (step B3).

In this case, the hiragana "a" standard pattern is expanded and converted in the vertical direction so that it becomes a double-height vertical pattern. Then, according to the input position of the writing frame, the obtained character pattern is developed at the corresponding position on the display data buffer 33, here the position of the writing frame 51a, and displayed at the corresponding display position of the display device 13 (step B5) With the above, the processing from the handwriting input to the display of the recognition result is completed.

Accordingly, the input process of the coordinate value series for the handwriting input by the input unit 21 and the subsequent character modification by the recognition data processing unit 22, the recognition processing unit 23, the display data processing unit 24, and the display processing unit 25 are included. Since the character recognition process and the recognition result display process are performed in parallel, the writer can continuously perform handwriting input while the recognition process and the display process are performed.

In the above-described embodiment, an example is shown in which a character is double-horizontally or vertically double-doublet-corrected to an input character by performing a handwriting input so as to extend out of the writing frame. The type of employment is not limited to the above embodiment, and various other combinations can be selected and set. As described above, the present invention is not limited to the above-described embodiments, and various modifications can be carried out without departing from the scope of the invention.

[0043]

As described above, according to the present invention, there is provided a character recognition apparatus and method capable of modifying an input character without performing a special operation in response to the handwriting of the character. can do.

[Brief description of drawings]

FIG. 1 is a block diagram showing a schematic circuit configuration according to an embodiment of the present invention.

FIG. 2 is a diagram schematically showing a schematic device configuration according to the embodiment.

FIG. 3 is a diagram exemplifying a writing frame displayed on the display device screen of FIG.

FIG. 4 is a diagram illustrating a writing state of characters in the writing frame in FIG.

5 is a diagram illustrating a writing state of characters in the writing frame of FIG.

FIG. 6 is a flowchart showing the processing contents mainly by the recognition data processing unit in FIG.

FIG. 7 is a flowchart mainly showing the processing contents of the display processing unit in FIG.

[Explanation of symbols]

11 ... Coordinate input device, 12 ... Control device, 13 ... Display device, 21 ...
Input unit, 22 ... Recognition data processing unit, 23 ... Recognition processing unit, 24 ...
Display data processing unit, 25 ... Display processing unit, 31 ... Position data buffer, 32 ... Recognition data buffer, 33 ... Display data buffer, 34 ... Character data buffer, 41 ... Character recognition dictionary, 42
... Character pattern dictionary, 51a to 51f ... Writing frame.

─────────────────────────────────────────────────── ─── Continuation of the front page (51) Int.Cl. ⁶ Identification code Internal reference number FI Technical display location G06K 7/00 D 9069-5L 9/20 340 C (72) Inventor Satoshi Kawamura Kawasaki City, Kanagawa Prefecture 70 Yanagimachi, Saiwai-ku, Toshiba Yanagimachi Factory

Claims (2)

[Claims] 1. A coordinate input means having a writing frame for obtaining a coordinate value sequence of characters written and input to the writing frame, and a coordinate value sequence obtained by the coordinate input means is a fixed amount of the writing frame. The above-mentioned coordinates are determined by referring to the above-mentioned recognition dictionary based on the coordinate value sequence obtained by the above coordinate input sequence A character recognition means for recognizing a character entered in the writing frame of the input means; and a character modification means for adding character modification information to the character recognized by the character recognition means according to the judgment result of the judgment means and outputting the character. A character recognizing device, comprising: display means for displaying the character information output by the character modifying means in a position corresponding to the writing frame of the coordinate input means in a character pattern according to the presence or absence of character employment information. 2. A coordinate input process for obtaining a coordinate value sequence when a character is handwritten and input to a preformed writing frame, and a coordinate value sequence obtained by this coordinate input process is a fixed amount of the writing frame. Based on the judgment process for judging whether or not it is protruding, and the coordinate value sequence obtained by the coordinate input process, the recognition dictionary in which the character information corresponding to the coordinate value sequence is registered in advance is referred to A character recognition process for recognizing an input character, a character modification process for adding character modification information to the character recognized by the character recognition process according to the judgment result of the judgment process, and outputting the character modification means. A character recognition method comprising: displaying character information in a character pattern according to the presence or absence of character employment information at a position corresponding to the writing frame on a display screen.